Dr. Peter Shea
Does Online Learning Inhibit or Support Community College Student Success?
Using a nationally representative sample (The Beginning Postsecondary Student Survey, BPS 04/09), this study examined the associations between enrollment in credit-bearing distance education courses and degree attainment. We sought to determine whether US students enrolled in online/distance education courses during their first year of study at a community college tend to complete a degree (certificate, associate, or bachelor’s) at significantly lower rates than those who were not enrolled in such courses or programs. Unlike previous researchers, our findings indicated that students who take some of their early courses online or at a distance have a significantly better chance of attaining a community college credential than do their classroom-only counterparts. Implications for policy, practice, and theory related to student attrition, persistence, and success will be discussed.
presentation at the 15th annual SLN SOLsummit 2014 February 26, 2014
http://slnsolsummit2014.edublogs.org
Peter Shea: Does Online Learning Inhibit or Support Community College Student Success?
1. Dr. Peter Shea, Senior Researcher, SLN
University at Albany, SUNY
Dr. Temi Bidjerano
Furman University
2.
Open SUNY – Open Journal
Editor for the Journal of Asynchronous
Learning Networks (JALN)
Reinvigorating the journal
You are invited!
http://jaln.sloanconsortium.org/
Register to be considered as a reviewer
http://jaln.sloanconsortium.org/index.php/i
ndex/user/register
7. What is “access”?
“Full” benefits of higher education
Role of OLL in improving access to
benefits
Previous theoretical and empirical work
New national study
› Question, methods, results, discussion
8. How do we measure access to HE?
Access to what?
What is the role of flexibility and
convenience of online learning?
Has online learning improved access?
How do we know?
Not as obvious as you might think…
9.
10.
Many (not all) of these benefits are
contingent on completing a degree, not
simply going to college…
11.
12. US not producing sufficient numbers of
college graduates
Losing competitive advantage in the global
economy (Hebel, 2006; Kelderman, 2013)
Community Colleges:
› Six year national completion rates less than 20%
› Justifiably or not, community colleges are target
of criticism
› But community colleges crucial to supporting US
economy (College Board, 2008).
13.
14. Spending on community college students
has poor return on degree attainment
Delta project (Kirshtein & Wellman, 2012)
› “…half of instructional spending in community
colleges goes to students (and credits) that do
not attach to a degree or certificate” (p. 16)
› Raises questions about efforts to increase access
to community college students
› Is it expensive and inefficient relative to benefits?
15.
16.
Dramatic growth: 5.5-7M+ online enrollments in US
Most of them community college students but…
Does online learning merely increase the pool of a
costly population of higher education learners who
do not complete?
Crazy, right? Strong evidence that this may be the
case…
But first, what theories inform research on degree
completion?
20. Other Models: Falcone, 2012
Assumes
these as
“given”
Multivariate
Outcomes
Assumes
these as
“given”
21.
What solid evidence do we have about
online learning, degree completion,
persistence, transfer etc?
22.
23.
Online learning outcomes worse
› Smith Jaggars & Xu, 2010; Xu & Smith
Jaggars, 2011
Data: N=24,000 in 23 institutions in Virginia
Community College System
More failing/withdrawing from online
Online students less likely to return
Students w higher proportion of credits
online less likely to attain credential /transfer
to 4 year institution
24. Data: N=51,000 in 34 institutions in
Washington State Community College
System (Xu & Smith Jaggars, 2011)
Students with better preparation more
likely to enroll in online courses…
But more likely to fail/withdraw
Students who took more online courses
less likely to complete degree or transfer
25. Best evidence does not support strategy
of increasing access via online ed
Produces more college students but
fewer with vital college credentials
OLL less efficient/effective in goal of
producing more college grads in US
Questionable results re. “meaningful”
access
26. Our study uses a national, rather than
state samples of community college
students
Question: Does a national sample yield
same results?
27.
28. Initial analysis of this NCES data revealed
interesting patterns
Community College student seemed to
do better with some distance courses
Even better when the courses were
online …
29. Cumulative persistence and attainment anywhere 2008-09
Attained
Attained
Attained
No degree,
bachelor's degree associate's degree
certificate
still enrolled
No degree,
left without
return
Total
(%)
(%)
(%)
(%)
(%)
30.7
9.3
9.4
15
35.5
100%
8.9
13.5
9.6
8
15.1
14.3
35.2
38
100%
100%
0.37
0.38
0.55
0.65
Estimates
Total
Distance education 2004: Took courses
No
31.2
Yes
26.3
Standard Error (BRR)
Total
0.56
52% higher associates degree
attainment?
30. Cumulative persistence and attainment anywhere 2008-09
Attained
Attained
bachelor's associate's
degree
degree
Attained No degree, No degree,
certificate
still
left
enrolled
without
return
Total
(%)
(%)
(%)
(%)
(%)
30.7
9.3
9.4
15
35.5
100%
Distance education used 2004: Internet
No
31.2
Yes
25.5
8.9
14.1
9.6
7.3
15
15.1
35.3
38
100%
100%
0.37
0.38
0.55
0.65
Estimates
Total
Standard Error (BRR)
Total
0.56
58% higher associates degree
attainment?
31. Does not account for other factors
Possible that initial differences among
these two groups accounts for the higher
degree completion rate
Took steps to control for initial differences
between students in two groups
32. Degree attainment modeled as a
function of enrollment in online/distance
education courses at a community
college
Controlling for a range of background
characteristics
33. (a) person variables: gender, age, race, risk index of
dropping out (NCES derived), disability status, remedial
coursework eligibility, traditional high school diploma or
not (GED, certificate of completion, homeschooled), type
of high school (public, private, other), total amount of
loans during the first year etc.
(b) family variables: gross adjusted family income, parents‟
highest level of education, family size, siblings in college
before respondent, parents taking college courses,
parents‟ place of birth (U.S. vs. not);
(c) institutional variables: distance from home, historically
black institution, Hispanic serving institution, accreditation,
in-state institution, size of enrollment, percent of student
body receiving federal grants, and type of institution (rural,
suburban, urban or other).
34. Matching attempts to mimic
randomization in experimental design
Create a sample that received the
treatment (DE/Online Ed)
Comparable on all 40 observed
covariates to a sample that did not
receive the treatment
Look at differences between these now
matched samples
35.
Data for BPS 04/09 collected from 16,100
first time beginning students at three points
in time (in spring of 2004, 2006, 2009)
Participants in BPS represent target
population of approximately 4,000,000 first
time beginning post secondary students.
43.1% were first enrolled in two-year
institutions offering associate‟s or certificate
degree.
36. Net of 40 factors: More likely to take distance/online courses
› Female students (p<.001), older students (p<.001), students from
larger families (p<.05), students with a higher amount of
institutional aid (p<.01) and loans (p<.05)
› Students whose residence was at a greater distance from the
institution
› At greater risk of not completing a degree were somewhat more
likely to be enrolled in distance education courses (p<.10).
Net of 40 factors: Less likely to take distance/online courses
› African –American students as compared to white students
(p<.001)
› Students who had indicated that location (p<.05) represents a
reason for attending a particular institution.
37.
Positive Effects:
› Number of months of full time enrollment increase the chances
of degree attainment (B=.062, p<.001).
› Students whose initial goal was to earn a certificate were about
three times more likely to graduate (B = 1.168, p<.001)
Neutral Effects
› No differences in the odds of degree completion between
students with plans for a bachelor and those with a goal to earn
an associate degree (B =.184, p>.05).
› Number of institutions attended does not decrease the odds for
degree (B = .072, n.s.).
Negative Effects
› The number of periods of interruptions in continuous enrollment
had a negative effect on the likelihood of degree attainment
(B=-.232, p<.01)
38.
Distance Education Students:
At higher risk for dropout
Didn‟t attend private high schools
Maybe not better prepared?
Despite this…
39.
40. Net of 40 other factors DE/online learners
were 1.25 times as likely to attain any
credential
When credential goal was certificate
(rather than BA) DE/online learners were
3.22 times as likely to succeed
41. Net of other differences the odds of
graduating increase for students with
early distance education coursework
Participation in online/distance learning
does not appear to impede degree
completion – perhaps facilitates it.
44. Peter Shea, PhD
Associate Professor
Educational Theory and Practice &
College of Computing and Information
University at Albany, State University of
New York
pshea@albany.edu
45.
Several notable findings:
› Unlike previous researchers (Xu & Smith
Jaggars, 2010; Smith-Jaggars and Xu, 2011)
› We did not find that students who
participated in online/distance education
were better-prepared academically
› The online/distance students were about the
same or maybe less well prepared
46. Despite potential initial disadvantage did
not replicate findings that online/distance
community college students were less likely
to complete a college credential.
National level data yields the opposite
conclusion.
Evidence suggests that early participation
in online learning and distance education
predicts higher rates of community college
degree attainment.
47.
Women overrepresented in DE/Online courses
DE/Online students more likely to receive
financial aid and have loans
An artifact of higher levels of enrollment in
private, for profit institutions?
For profits have demonstrated capacity to
ensure DE students get maximum financial aid
relative to public institutions (Clayton, 2011)
Federal student loan data indicates students at
for-profit institutions borrow more (and default
more frequently) than those at public
institutions (US Dept of Education, 2010)
Editor's Notes
US is not producing sufficient numbers of college graduates Losing competitive advantage in the global economy (Hebel, 2006; Kelderman, 2013)Community Colleges:Six year national completion rates less than 20% Justifiably or not, community colleges are target of a great deal of criticismBut community colleges are crucial to supporting the US economy (College Board, 2008).
In 1985, Bean and Metzner developed a conceptual model, based on the Tinto model for non-traditional students. Learner characteristics for this model included age, enrollment status, residence, educational goals, high school performance, ethnicity, and gender. Academic variables included study habits, advising, absenteeism, certainty of major and course availability leading to academic outcomes (GPA). In place of the social supports of a campus, Bean and Metzner saw environmental factors such as finances, family responsibilities, employment responsibilities, and especially outside encouragement (from work and home) as being major determinants of psychological outcomes – utility of courses, satisfaction, goal commitment, and stress --which they in turn saw as a major determinant of the decision to drop-out. So for Bean & Metzner, <ADVANCE> learner characteristics, environmental variables and psychological outcomes (especially course satisfaction and perceived utility) were at the core of retention for non traditional students
Data: 24,000 students in 23 institutions in the Virginia Community College SystemOnline learning outcomes at community college worse than comparable classroom based outcomesSmith Jaggars & Xu, 2010; Xu & Smith Jaggars, 2011:Greater likelihood of failing or withdrawing from online courses than face-to-face coursesStudents who took online coursework in early semesters less likely to return in following semestersStudents w higher proportion of credits online less likely to attain a credential or transfer to a 4 year institution.
Data: N=51,000 in 34 institutions in Washington State Community College System (Xu & Smith Jaggars, 2011) Students with better educational preparation more likely to enroll in online courses… But significantly more likely to fail or withdraw from these courses than students who took face-to-face classesStudents who took more online courses less likely to complete a degree or transfer to a four-year college than those who took fewer online courses
Best evidence does not support strategy of increasing access to community college through online education. Given high costs associated with low completion rates online learning appears to produces more college students but fewer students with vital college credentials upon completion Online learning is less efficient and effective in achieving the goals of producing more college graduates in the US
Current study: build upon and add to the literature on the impacts of participation in online and distance education on community college degree completion Uses a national, rather than state samples of community college studentsDoes a national sample yield same results?
Does not account for other factors that contribute to higher success rates among online students Possible that initial differences among these two groups accounts for the higher degree completion rate. Took steps to control for initial differences between students in two groups
Matching in the propensity score approach attempts to mimic randomization in experimental designCreate a sample that received the treatment (DE) that is comparable on all observed covariates to a sample that did not receive the treatmentLook at differences between these matched samples
Higher Risk for dropout: Using an NCES derived risk category for drop out we found an overrepresentation of higher risk drop out among distance students on six of the seven variables that make up that category. Didn’t Attend Private High Schools: We also found that students who went to private high schools were underrepresented among those who took online and distance courses. Not better prepared? We therefore conclude that online and distance students appear to be no better academically prepared and possibly somewhat less prepared and/or less likely to graduate than students who did not take distance or online courses.
Net of the influence of the forty person, family, and institutional covariates for which we controlled via propensity matching, students who had taken a distance or online course early in their academic careers were 1.25 times as likely to attain any degree as compared to those who had not taken a such a course.
Despite potential initial disadvantage did not replicate findings that online/distance community college students were less likely to complete a college credential. National level data yields the opposite conclusion. Evidence suggests that early participation in online learning and distance education predicts higher rates of community college degree attainment.
Women overrepresented in DE/Online courses DE/Online students more likely to receive financial aid and have loansAn artifact of higher levels of enrollment in private, for profit institutions?For profits have demonstrated capacity to ensure DE students get maximum financial aid relative to public institutions (Clayton, 2011)Federal student loan data indicates students at for-profit institutions borrow more (and default more frequently) than those at public institutions (US Dept of Education, 2010)